In business, almost all executive decisions (headcount and budget allocation, which unproven products to push ahead with aggressively, translating forecasts for macroeconomic risks into business-specific policies, who to promote to other executive level positions, etc.) are made with substantial uncertainty. Or to put it another way, any executive-level decision-maker would be paralyzed without strong priors. This is especially true in fast-changing or competitive markets, where the only way to collect more evidence without direct risk is to let your competitors jump in the water first.
In other words, the kind of certainty we hold out for (often vainly) in science is almost unknown in many aspects of business, and the most critical decisions are often the most uncertain.
It’s very “Black Swan” (in the sense of Taleb’s whole, not just tail risk).
any executive-level decision-maker would be paralyzed without strong prior
I don’t think that’s necessarily true, just having a high risk tolerance works as well. I also think you underestimate the amount of evidence present—e.g. in most organizations the next-year budget is a variation on the previous year’s budget.
the kind of certainty we hold out for (often vainly) in science is almost unknown in many aspects of business
Yes, of course. That’s why, for example, risk management is an important part of doing business but is not normally a big part of doing science...
Risk tolerance is a good, possibly more correct, way of looking at it. Actually most executives probably have a mixture of risk tolerance and strong priors.
Some businesses can get away with only relatively low-risk, safe decisions and focus on efficient operations. However, I think the majority of businesses, especially newer and growing ones, can’t get away with this consistently or for a long time. And most businesses simply don’t have that long a life, period.
Setting a budget based off last years’ when your revenue is growing 50%+ YoY won’t work well.
What I was thinking of more specifically is that something like setting a budget can be defined as a rigorous optimization problem, but with highly uncertain parameters (marginal return on investment from various units of the business). Any decision made implies a combination of prior over those values and risk tolerance.
Any decision made implies a combination of prior over those values and risk tolerance.
If you treat budgeting as an optimization problem, you need forecasts, not priors.
I would also suspect that real-life business budgets will be hard to set as “rigorous optimization problems” because in reality you have discontinuities, nonlinear responses, and all kinds of funky dependencies between different parts of the budget.
I don’t think you understand what the term means. It’s unknown unknowns and not known unknowns. Whether or not an unproven product will succeed is a question about a known unknown.
This is especially true in fast-changing or competitive markets, where the only way to collect more evidence without direct risk is to let your competitors jump in the water first.
I don’t think that’s true. There are various forms of doing market research that simply involve money but not additional risk.
I use “Black Swan” in the context of the whole book. That is, we build narratives after-the-fact to explain correct priors as skill and judgment. Also, the greater impact of more uncertain decisions, in a way that ties uncertainty to the impact, is exactly the nature of unknown-unknown black swans (which I’d say the launching of a substantially new product category fits into, in a mild form. The iPod/iTunes was not a black swan for Apple, though they took considerable risks with it. It was a black swan for the music industry.).
Market research is better than nothing, but still has many problems. Most of it wouldn’t pass peer review, and we know peer review makes plenty of mistakes. So when taking it into account, decision-makers must apply strong priors.
And on the occasions that market research really is that good, it’s a no-brainer; your competitors will do it too.
I use “Black Swan” in the context of the whole book
Please don’t take terminology with fairly precise meaning and use it idiosyncratically. At best, you unnecessary increase your inferential distance. At worst, you dilute the term so that it increases everyone’s inferential distance.
In business, almost all executive decisions (headcount and budget allocation, which unproven products to push ahead with aggressively, translating forecasts for macroeconomic risks into business-specific policies, who to promote to other executive level positions, etc.) are made with substantial uncertainty. Or to put it another way, any executive-level decision-maker would be paralyzed without strong priors. This is especially true in fast-changing or competitive markets, where the only way to collect more evidence without direct risk is to let your competitors jump in the water first.
In other words, the kind of certainty we hold out for (often vainly) in science is almost unknown in many aspects of business, and the most critical decisions are often the most uncertain.
It’s very “Black Swan” (in the sense of Taleb’s whole, not just tail risk).
Thoughts?
I don’t think that’s necessarily true, just having a high risk tolerance works as well. I also think you underestimate the amount of evidence present—e.g. in most organizations the next-year budget is a variation on the previous year’s budget.
Yes, of course. That’s why, for example, risk management is an important part of doing business but is not normally a big part of doing science...
Risk tolerance is a good, possibly more correct, way of looking at it. Actually most executives probably have a mixture of risk tolerance and strong priors.
Some businesses can get away with only relatively low-risk, safe decisions and focus on efficient operations. However, I think the majority of businesses, especially newer and growing ones, can’t get away with this consistently or for a long time. And most businesses simply don’t have that long a life, period.
Setting a budget based off last years’ when your revenue is growing 50%+ YoY won’t work well.
What I was thinking of more specifically is that something like setting a budget can be defined as a rigorous optimization problem, but with highly uncertain parameters (marginal return on investment from various units of the business). Any decision made implies a combination of prior over those values and risk tolerance.
If you treat budgeting as an optimization problem, you need forecasts, not priors.
I would also suspect that real-life business budgets will be hard to set as “rigorous optimization problems” because in reality you have discontinuities, nonlinear responses, and all kinds of funky dependencies between different parts of the budget.
I don’t think you understand what the term means. It’s unknown unknowns and not known unknowns. Whether or not an unproven product will succeed is a question about a known unknown.
I don’t think that’s true. There are various forms of doing market research that simply involve money but not additional risk.
I use “Black Swan” in the context of the whole book. That is, we build narratives after-the-fact to explain correct priors as skill and judgment. Also, the greater impact of more uncertain decisions, in a way that ties uncertainty to the impact, is exactly the nature of unknown-unknown black swans (which I’d say the launching of a substantially new product category fits into, in a mild form. The iPod/iTunes was not a black swan for Apple, though they took considerable risks with it. It was a black swan for the music industry.).
Market research is better than nothing, but still has many problems. Most of it wouldn’t pass peer review, and we know peer review makes plenty of mistakes. So when taking it into account, decision-makers must apply strong priors.
And on the occasions that market research really is that good, it’s a no-brainer; your competitors will do it too.
Please don’t take terminology with fairly precise meaning and use it idiosyncratically. At best, you unnecessary increase your inferential distance. At worst, you dilute the term so that it increases everyone’s inferential distance.
Edited for clarity. Thought terms get diluted all the time.
Maybe “Talebian” would be more appropriate.